CDT Files Comments to White House Office of Science and Technology Policy Regarding the Federal Evidence Agenda on Disability Equity

Today, the Center for Democracy & Technology (CDT) submitted a comment to the White House Office of Science and Technology (OSTP) in response to its request for information on building the Federal Evidence Agenda on Disability Equity. The Agenda aims to improve the federal government’s ability to make data-informed policy decisions and to ensure those decisions advance equity for people with disabilities. 

CDT commends OSTP for requesting public input on this subject, and for its commitment to working towards real equity for people with disabilities. As experts in technology policy, CDT has long recognized how underinclusive or non-inclusive data sets can contribute to algorithmic bias, and how those biased algorithmic systems can impact marginalized communities, particularly people with disabilities. 

Algorithmic systems are trained on datasets, where the system learns patterns included in the dataset, and then uses those patterns to create new outputs, whether those be a credit score, a risk score, or novel text. When datasets that are used to build and train algorithmic models are not fully representative of people with disabilities, they can amplify algorithmic bias and its impacts. To minimize these effects, it is important to train algorithms on datasets that are representative of people with disabilities, and to use those datasets to evaluate and mitigate an algorithmic system’s risks to disabled people. The existence of a thoughtful federal strategy on disability equity in data will be a helpful tool in ensuring that data is gathered in ways that mitigate discrimination in many contexts, including technology-facilitated disability discrimination.

CDT responded to questions asked by OSTP regarding describing disparities, informing data collection and public access, and privacy, security and civil rights. In particular, we highlighted the following: 

• there are at least four different models of defining disability (legal, medical, social, and identity/demographic), and the existence of varying models of disability can influence the outcomes of data gathering and collecting processes. 

• barriers such as the existence of disability-related stigma and the disproportionate incarceration and institutionalization of disabled people can make accurate data collection for people with disabilities more difficult. 

• agencies should incorporate various privacy protection mechanisms into their data collection practices when gathering disability-related data, including using encryption, de-identification, and aggregation when possible. In addition, agencies should only retain data for as long as is necessary. 

We look forward to continuing to collaborate with OSTP in creating more equitable data collection practices for people with disabilities, and to the publication of the Federal Evidence Agenda. 

Read the full comments here.